Pub Date : 2022-03-25DOI: 10.1109/ICCAE55086.2022.9762439
M. Nabil, Dalia M. Mahfouz, Omar M. Shehata
Research in the field of robotics and industries is heavily investing in human-robot interaction. Collaborative robots aim to bridge the gap and bring humans and robots closer together in industrial settings by taking down physical barriers and sharing workspaces. This paper describes an integrated system architecture for modeling and controlling robots for human-collaborative behavior, while employing safety assurance mechanisms. This is performed by integrating motion tracking and compliance algorithms implemented using linear Proportional-Integral (PI) and Admittance controllers, respectively. The controllers are applied on the robot’s kinematic and dynamic models, alongside with force/torque and proximity sensors. The developed architecture is verified through a series of simulations applied using MATLAB/Simulink environment and Simscape Multibody visualization method.
{"title":"Development and Evaluation of a Control Architecture for Human-Collaborative Robotic Manipulator in Industrial Application","authors":"M. Nabil, Dalia M. Mahfouz, Omar M. Shehata","doi":"10.1109/ICCAE55086.2022.9762439","DOIUrl":"https://doi.org/10.1109/ICCAE55086.2022.9762439","url":null,"abstract":"Research in the field of robotics and industries is heavily investing in human-robot interaction. Collaborative robots aim to bridge the gap and bring humans and robots closer together in industrial settings by taking down physical barriers and sharing workspaces. This paper describes an integrated system architecture for modeling and controlling robots for human-collaborative behavior, while employing safety assurance mechanisms. This is performed by integrating motion tracking and compliance algorithms implemented using linear Proportional-Integral (PI) and Admittance controllers, respectively. The controllers are applied on the robot’s kinematic and dynamic models, alongside with force/torque and proximity sensors. The developed architecture is verified through a series of simulations applied using MATLAB/Simulink environment and Simscape Multibody visualization method.","PeriodicalId":294641,"journal":{"name":"2022 14th International Conference on Computer and Automation Engineering (ICCAE)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115003524","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-25DOI: 10.1109/ICCAE55086.2022.9762421
Mohsen Eskandari, Hailong Huang, A. Savkin, Wei Ni
The excellent 3D mobility has made aerial drones (ADs) good candidates to improve the coverage of wireless communication networks. However, the effectiveness/efficiency of an AD (as an aerial active relay) is questionable for holding a continuous line-of-sight (LoS) link with a moving node (MN) in fifth-generation (5G) millimeter-wave (mmWave) wireless networks due to the signal propagations/blockage issues. Nevertheless, intelligent reflecting surface (IRS) technology has been introduced as a useful and energy-efficient method to improve the spectrum efficiency of 5G networks. In this paper, an AD equipped with IRS is proposed and autonomously navigated to maintain a continuous wireless link with an MN. The MN can be a vehicle moving through a dense urban environment. Simulation results corroborate the effectiveness of the method.
{"title":"Autonomous Guidance of an Aerial Drone for Maintaining an Effective Wireless Communication Link with a Moving Node Using an Intelligent Reflecting Surface","authors":"Mohsen Eskandari, Hailong Huang, A. Savkin, Wei Ni","doi":"10.1109/ICCAE55086.2022.9762421","DOIUrl":"https://doi.org/10.1109/ICCAE55086.2022.9762421","url":null,"abstract":"The excellent 3D mobility has made aerial drones (ADs) good candidates to improve the coverage of wireless communication networks. However, the effectiveness/efficiency of an AD (as an aerial active relay) is questionable for holding a continuous line-of-sight (LoS) link with a moving node (MN) in fifth-generation (5G) millimeter-wave (mmWave) wireless networks due to the signal propagations/blockage issues. Nevertheless, intelligent reflecting surface (IRS) technology has been introduced as a useful and energy-efficient method to improve the spectrum efficiency of 5G networks. In this paper, an AD equipped with IRS is proposed and autonomously navigated to maintain a continuous wireless link with an MN. The MN can be a vehicle moving through a dense urban environment. Simulation results corroborate the effectiveness of the method.","PeriodicalId":294641,"journal":{"name":"2022 14th International Conference on Computer and Automation Engineering (ICCAE)","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114945645","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-25DOI: 10.1109/ICCAE55086.2022.9762432
Yingming Wang, Qian Zhang, Wei Feng
Shadow is an important clue to the relation between the scene’s structure and lighting condition. However, most photometric stereo algorithms treat shadows as outliers and thus lost this information. In this work, we propose a shadow-aware photometric stereo network that explicitly takes advantage of shadow information. We introduce a shadow estimation model to detect cast shadows and design a reconstruction loss based on the estimated shadow map. On the one hand, adding the shadow information in the reconstruction loss can effectively reduce the influence of scene shadow to normal estimation. On the other hand, the proposed shadow estimation model solves the bas-relief ambiguity problem in uncalibrated photometric stereo. Experiments show the superiority of our SAPS-Net against other uncalibrated photometric stereo algorithms. Besides, the proposed reconstruction loss makes it possible for SAPS-Net to be optimized on real-world data by fine-tuning itself in a self-supervised way, making our method more practical.
{"title":"Shadow-aware Uncalibrated Photometric Stereo Network","authors":"Yingming Wang, Qian Zhang, Wei Feng","doi":"10.1109/ICCAE55086.2022.9762432","DOIUrl":"https://doi.org/10.1109/ICCAE55086.2022.9762432","url":null,"abstract":"Shadow is an important clue to the relation between the scene’s structure and lighting condition. However, most photometric stereo algorithms treat shadows as outliers and thus lost this information. In this work, we propose a shadow-aware photometric stereo network that explicitly takes advantage of shadow information. We introduce a shadow estimation model to detect cast shadows and design a reconstruction loss based on the estimated shadow map. On the one hand, adding the shadow information in the reconstruction loss can effectively reduce the influence of scene shadow to normal estimation. On the other hand, the proposed shadow estimation model solves the bas-relief ambiguity problem in uncalibrated photometric stereo. Experiments show the superiority of our SAPS-Net against other uncalibrated photometric stereo algorithms. Besides, the proposed reconstruction loss makes it possible for SAPS-Net to be optimized on real-world data by fine-tuning itself in a self-supervised way, making our method more practical.","PeriodicalId":294641,"journal":{"name":"2022 14th International Conference on Computer and Automation Engineering (ICCAE)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114575863","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-25DOI: 10.1109/ICCAE55086.2022.9762429
V. O. K. Tan, Thu Nguyen
Blockchain is one of the leading 4.0 technologies because of its potential applications in many fields, such as real estate. Real estate is an old market and little changes. Therefore, buying and selling real estate brings many risks because it is difficult to find and verify information. Real estate transactions are complex, time-consuming and have high intermediary fees. However, Blockchain technology has opened up many ways to change this. Therefore, we study Blockchain technology and propose Real Estate Transaction Trace (RETT) system model to change and improve the way real estate transactions work in Vietnam. The RETT system model is capable of managing and tracking the entire process of real estate transactions of all participants (owners, buyers, sellers, notaries). The government or real estate investor consortium can easily manage and trace all the purchase or lease transaction history of a property (land, town-house, apartment). We designed Blockchain network architecture, distributed ledger and smart contract. As a result, our method can digitize assets on the Blockchain, store decentralized transaction history, enable encryption, and direct transactions between sellers and buyers. In addition, our system model can reduce the problem of data explosion, execute multiple transactions at the same time, prevent data tampering and sensitive information disclosure. We have built a prototype of this system model based on the Ethereum Blockchain platform. Thereby, we have proven the effectiveness of RETT system model and practical applicability through experimental transactions. Our method improves transparency, eliminates intermediaries, saves costs and increases mutual trust in real estate transactions. Our system model can manage real estate transactions comprehensively and can be able to overcome the weaknesses of the current real estate transaction management system according to the Client-Server model. Therefore, our system model can be deployed in the smart city to manage and trace the entire real estate transaction.
{"title":"The Real Estate Transaction Trace System Model Based on Ethereum Blockchain Platform","authors":"V. O. K. Tan, Thu Nguyen","doi":"10.1109/ICCAE55086.2022.9762429","DOIUrl":"https://doi.org/10.1109/ICCAE55086.2022.9762429","url":null,"abstract":"Blockchain is one of the leading 4.0 technologies because of its potential applications in many fields, such as real estate. Real estate is an old market and little changes. Therefore, buying and selling real estate brings many risks because it is difficult to find and verify information. Real estate transactions are complex, time-consuming and have high intermediary fees. However, Blockchain technology has opened up many ways to change this. Therefore, we study Blockchain technology and propose Real Estate Transaction Trace (RETT) system model to change and improve the way real estate transactions work in Vietnam. The RETT system model is capable of managing and tracking the entire process of real estate transactions of all participants (owners, buyers, sellers, notaries). The government or real estate investor consortium can easily manage and trace all the purchase or lease transaction history of a property (land, town-house, apartment). We designed Blockchain network architecture, distributed ledger and smart contract. As a result, our method can digitize assets on the Blockchain, store decentralized transaction history, enable encryption, and direct transactions between sellers and buyers. In addition, our system model can reduce the problem of data explosion, execute multiple transactions at the same time, prevent data tampering and sensitive information disclosure. We have built a prototype of this system model based on the Ethereum Blockchain platform. Thereby, we have proven the effectiveness of RETT system model and practical applicability through experimental transactions. Our method improves transparency, eliminates intermediaries, saves costs and increases mutual trust in real estate transactions. Our system model can manage real estate transactions comprehensively and can be able to overcome the weaknesses of the current real estate transaction management system according to the Client-Server model. Therefore, our system model can be deployed in the smart city to manage and trace the entire real estate transaction.","PeriodicalId":294641,"journal":{"name":"2022 14th International Conference on Computer and Automation Engineering (ICCAE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130568332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-25DOI: 10.1109/ICCAE55086.2022.9762431
J. Villaverde, Ralph Emerson D. De Belen, Anne Bernadette M. Cañezo
Phosphate levels in ambient waters have traditionally been monitored manually using sampling techniques in which samples are collected in various designated areas or locations and brought to a laboratory for analysis and experimentation using traditional methods. The study's primary goal is to create a microfluidic system that recognizes and detects phosphate in ecological waters. The sample water is mixed with a vanadate-molybdate reagent to generate a yellow color. The absorbance of this resulting solution is analyzed by a spectral sensor integrated into the microfluidic chip to detect phosphate in the sample water. A two-tailed sample t-test was used to evaluate and was categorized to the alternative hypothesis if the resulting data are far from the expected values. After gathering the results and performing the statistical test, the result falls under the confidence level of 0.0521. Therefore, there is no significant difference between the manual collection and the microfluidic system.
{"title":"Microfluidic Device for Phosphate Detection on Environmental Waters","authors":"J. Villaverde, Ralph Emerson D. De Belen, Anne Bernadette M. Cañezo","doi":"10.1109/ICCAE55086.2022.9762431","DOIUrl":"https://doi.org/10.1109/ICCAE55086.2022.9762431","url":null,"abstract":"Phosphate levels in ambient waters have traditionally been monitored manually using sampling techniques in which samples are collected in various designated areas or locations and brought to a laboratory for analysis and experimentation using traditional methods. The study's primary goal is to create a microfluidic system that recognizes and detects phosphate in ecological waters. The sample water is mixed with a vanadate-molybdate reagent to generate a yellow color. The absorbance of this resulting solution is analyzed by a spectral sensor integrated into the microfluidic chip to detect phosphate in the sample water. A two-tailed sample t-test was used to evaluate and was categorized to the alternative hypothesis if the resulting data are far from the expected values. After gathering the results and performing the statistical test, the result falls under the confidence level of 0.0521. Therefore, there is no significant difference between the manual collection and the microfluidic system.","PeriodicalId":294641,"journal":{"name":"2022 14th International Conference on Computer and Automation Engineering (ICCAE)","volume":"25 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123558825","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-25DOI: 10.1109/ICCAE55086.2022.9762410
A. Yumang, Arianne I. Rojas, Clark Joshua R. Viray
Nowadays, as modern cryptocurrency machines operate faster and hotter, the fans are insufficient to cool constant CPU and GPU usage. The researchers built specialized Mining Rig Cases for these machines to withstand excessive heat. Aside from temperature, another environmental factor that affects the performance of Cryptocurrency machines is relative humidity. The researchers created an IoT-based Monitoring system to regulate and monitor the Cryptocurrency machines as excessive heat and humidity can decrease the lifespan of these machines. The Infrared camera used color mapping and Otsu’s Segmentation Algorithm for image processing. Otsu Segmentation Algorithm separates the background and foreground of the image while Color Mapping Algorithm converts colors of the image into temperature. The researchers also tested the power consumption and data sizes to determine the efficiency of the monitoring system. Using Convolutional Neural Network, the researchers trained 300 images to assess the state of the Cryptocurrency Mining rig. Additionally, a two-tailed T-Test will determine any significant difference between the two algorithms. Upon training the images, the results obtained show that the two-tail P-value of temperature and humidity is 0.35 and 0.2566, respectively, which affirms no significant difference. However, the power consumption and data size had a substantial difference with P-values of 0.0004 and 3E-06, respectively. Moreover, it shows that the application of Otsu reduces the data size and consumes more power due to deep learning.
{"title":"IoT-Based Monitoring of Temperature and Humidity Using Infrared Thermography for Cryptocurrency Mining Room","authors":"A. Yumang, Arianne I. Rojas, Clark Joshua R. Viray","doi":"10.1109/ICCAE55086.2022.9762410","DOIUrl":"https://doi.org/10.1109/ICCAE55086.2022.9762410","url":null,"abstract":"Nowadays, as modern cryptocurrency machines operate faster and hotter, the fans are insufficient to cool constant CPU and GPU usage. The researchers built specialized Mining Rig Cases for these machines to withstand excessive heat. Aside from temperature, another environmental factor that affects the performance of Cryptocurrency machines is relative humidity. The researchers created an IoT-based Monitoring system to regulate and monitor the Cryptocurrency machines as excessive heat and humidity can decrease the lifespan of these machines. The Infrared camera used color mapping and Otsu’s Segmentation Algorithm for image processing. Otsu Segmentation Algorithm separates the background and foreground of the image while Color Mapping Algorithm converts colors of the image into temperature. The researchers also tested the power consumption and data sizes to determine the efficiency of the monitoring system. Using Convolutional Neural Network, the researchers trained 300 images to assess the state of the Cryptocurrency Mining rig. Additionally, a two-tailed T-Test will determine any significant difference between the two algorithms. Upon training the images, the results obtained show that the two-tail P-value of temperature and humidity is 0.35 and 0.2566, respectively, which affirms no significant difference. However, the power consumption and data size had a substantial difference with P-values of 0.0004 and 3E-06, respectively. Moreover, it shows that the application of Otsu reduces the data size and consumes more power due to deep learning.","PeriodicalId":294641,"journal":{"name":"2022 14th International Conference on Computer and Automation Engineering (ICCAE)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129263590","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-25DOI: 10.1109/ICCAE55086.2022.9762447
Francis Jann Floresca, Christian Kyle Tobias, C. Ostia
Mechanical faults often occur in BLDC motors. These machines are essential to each respective industry. When a fault is not detected, it will cause the machine to stop functioning to its intended purpose. A mechanical fault diagnostic system using Naïve Bayes classifier with the DWT feature extraction method was proposed in this study. A single-level DWT was used to extract and decompose the recorded motor voltage signals, split into 7030, 70% for the training set, and 30% for the testing set. The accuracy in training the Naïve Bayes classifier is 97.2%. Using the trained model to the remaining test set resulted in an accuracy of 87.3% for detecting mechanical motor faults in a BLDC motor.
{"title":"Naïve Bayes Classification Technique for Brushless DC Motor Fault Diagnosis with Discrete Wavelet Transform Feature Extraction","authors":"Francis Jann Floresca, Christian Kyle Tobias, C. Ostia","doi":"10.1109/ICCAE55086.2022.9762447","DOIUrl":"https://doi.org/10.1109/ICCAE55086.2022.9762447","url":null,"abstract":"Mechanical faults often occur in BLDC motors. These machines are essential to each respective industry. When a fault is not detected, it will cause the machine to stop functioning to its intended purpose. A mechanical fault diagnostic system using Naïve Bayes classifier with the DWT feature extraction method was proposed in this study. A single-level DWT was used to extract and decompose the recorded motor voltage signals, split into 7030, 70% for the training set, and 30% for the testing set. The accuracy in training the Naïve Bayes classifier is 97.2%. Using the trained model to the remaining test set resulted in an accuracy of 87.3% for detecting mechanical motor faults in a BLDC motor.","PeriodicalId":294641,"journal":{"name":"2022 14th International Conference on Computer and Automation Engineering (ICCAE)","volume":"31 12","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113944563","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-25DOI: 10.1109/ICCAE55086.2022.9762450
Jian Zhang
Benefiting from the characteristics such as low-cost, excellent mobility and flexibility, rapid deployment, the aerial drones have been widely applied to wireless communication systems in both military and civilian domains. In particular,for the objects on the ground, UAVs could provide a high probability of line-of-sight (LoS) that enhances the capacity of the system. In this paper, we consider the application of covert surveillance for a solar-powered UAV on a suspicious mobile target on the ground while disguising its intention. A Q-learning based covert surveillance method is formulated to maximize the disguising metric and minimize the energy consumption while harvesting the solar energy while operation. Computer simulations demonstrate the performance of the proposed method, which outperforms the baseline method in both eavesdropping and disguising perspectives.
{"title":"A Covert Surveillance Strategy for a Solar-Powered UAV over Suspicious Mobile Target","authors":"Jian Zhang","doi":"10.1109/ICCAE55086.2022.9762450","DOIUrl":"https://doi.org/10.1109/ICCAE55086.2022.9762450","url":null,"abstract":"Benefiting from the characteristics such as low-cost, excellent mobility and flexibility, rapid deployment, the aerial drones have been widely applied to wireless communication systems in both military and civilian domains. In particular,for the objects on the ground, UAVs could provide a high probability of line-of-sight (LoS) that enhances the capacity of the system. In this paper, we consider the application of covert surveillance for a solar-powered UAV on a suspicious mobile target on the ground while disguising its intention. A Q-learning based covert surveillance method is formulated to maximize the disguising metric and minimize the energy consumption while harvesting the solar energy while operation. Computer simulations demonstrate the performance of the proposed method, which outperforms the baseline method in both eavesdropping and disguising perspectives.","PeriodicalId":294641,"journal":{"name":"2022 14th International Conference on Computer and Automation Engineering (ICCAE)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123404182","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-25DOI: 10.1109/ICCAE55086.2022.9762444
Boyu Li, Yongxin Wu, H. Sun, Zhiliang Cheng, Jiayi Liu
Robotic assembly is an important way to realize manufacturing intelligence. In the robotic assembly process, assembly sequence planning helps to improve the assembly efficiency. However, the existing research around assembly sequence planning always used hypothetical data or random data to get the optimal solution and it is not suitable for robotic assembly process because it does not consider the characteristics of the industrial robots. Aiming at this, in this paper, Unity-3D based simulation data driven robotic assembly planning using genetic algorithm is proposed in order to solve robotic assembly planning problem. The feasible assembly sequence is obtained by graph-based assembly precedence model. After that, the improved genetic algorithm is proposed. In the meanwhile, Unity3D-based simulation is utilized to obtain the simulation data for robotic assembly process. Based on a camera, the performance of the improved genetic algorithm is analyzed and compared with the traditional genetic algorithm. It shows the proposed method is suitable for robotic assembly sequence planning problem.
{"title":"Unity 3D-Based Simulation Data Driven Robotic Assembly Sequence Planning Using Genetic Algorithm","authors":"Boyu Li, Yongxin Wu, H. Sun, Zhiliang Cheng, Jiayi Liu","doi":"10.1109/ICCAE55086.2022.9762444","DOIUrl":"https://doi.org/10.1109/ICCAE55086.2022.9762444","url":null,"abstract":"Robotic assembly is an important way to realize manufacturing intelligence. In the robotic assembly process, assembly sequence planning helps to improve the assembly efficiency. However, the existing research around assembly sequence planning always used hypothetical data or random data to get the optimal solution and it is not suitable for robotic assembly process because it does not consider the characteristics of the industrial robots. Aiming at this, in this paper, Unity-3D based simulation data driven robotic assembly planning using genetic algorithm is proposed in order to solve robotic assembly planning problem. The feasible assembly sequence is obtained by graph-based assembly precedence model. After that, the improved genetic algorithm is proposed. In the meanwhile, Unity3D-based simulation is utilized to obtain the simulation data for robotic assembly process. Based on a camera, the performance of the improved genetic algorithm is analyzed and compared with the traditional genetic algorithm. It shows the proposed method is suitable for robotic assembly sequence planning problem.","PeriodicalId":294641,"journal":{"name":"2022 14th International Conference on Computer and Automation Engineering (ICCAE)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128171797","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-25DOI: 10.1109/ICCAE55086.2022.9762416
A. Yumang, Ma. Chloe M. Sta. Juana, Regina Liza C. Diloy
This study focuses on creating a system that detects and classifies defective fresh Excelsa beans. Mask R-CNN is a Convolutional Neural Network model that predicts classes and generates bounding boxes and segmentation masks for each class. The pre-trained model that will be used is called Detectron2 Mask R-CNN, and it will be trained in the Google Colab to learn the features of each Defective Fresh Excelsa bean, namely, Black Bean, Sour Bean, Cut Bean, and Insect Damaged Bean. To test the model’s accuracy in detecting and classifying the model, the researchers will use the Raspberry Pi 4 with a camera and take a picture of 40 fresh Excelsa beans. The system will automatically detect and classify the Fresh Excelsa bean. Then the output will be displayed in the Raspberry Pi LCD. With the gathered data, the model achieved an accuracy of 87.5%.
本研究的重点是创建一个检测和分类有缺陷的新鲜Excelsa豆的系统。Mask R-CNN是一种卷积神经网络模型，用于预测类并为每个类生成边界框和分割掩码。将使用的预训练模型称为Detectron2 Mask R-CNN，它将在Google Colab中进行训练，以学习每一种有缺陷的新鲜Excelsa豆的特征，即黑豆，酸豆，切豆和虫损豆。为了测试模型在检测和分类模型方面的准确性，研究人员将使用带有相机的树莓派4，并拍摄40颗新鲜Excelsa豆的照片。该系统将自动检测和分类新鲜Excelsa豆。然后在Raspberry Pi LCD上显示输出结果。利用收集到的数据，该模型的准确率达到87.5%。
{"title":"Detection and Classification of Defective Fresh Excelsa Beans Using Mask R-CNN Algorithm","authors":"A. Yumang, Ma. Chloe M. Sta. Juana, Regina Liza C. Diloy","doi":"10.1109/ICCAE55086.2022.9762416","DOIUrl":"https://doi.org/10.1109/ICCAE55086.2022.9762416","url":null,"abstract":"This study focuses on creating a system that detects and classifies defective fresh Excelsa beans. Mask R-CNN is a Convolutional Neural Network model that predicts classes and generates bounding boxes and segmentation masks for each class. The pre-trained model that will be used is called Detectron2 Mask R-CNN, and it will be trained in the Google Colab to learn the features of each Defective Fresh Excelsa bean, namely, Black Bean, Sour Bean, Cut Bean, and Insect Damaged Bean. To test the model’s accuracy in detecting and classifying the model, the researchers will use the Raspberry Pi 4 with a camera and take a picture of 40 fresh Excelsa beans. The system will automatically detect and classify the Fresh Excelsa bean. Then the output will be displayed in the Raspberry Pi LCD. With the gathered data, the model achieved an accuracy of 87.5%.","PeriodicalId":294641,"journal":{"name":"2022 14th International Conference on Computer and Automation Engineering (ICCAE)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134469193","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：